Columns in blended_scores_v2.tab:

question_id: variable in the original dataset
prob.unw: unweighted mean in the probability sample
conv.unw: unweighted mean in the convenience (nonprobability) sample
p.val.unw: p-value of the difference in unweighted means
prob.w: weighted mean in the probability sample
conv.w: weighted mean in the convenience sample
variable: variable description
scale_l: lower bound of the scale used to normalize each variable
scale_u: upper bound of the scale used to normalize each variable
reverse: indicates variable needs to be reverse coded